Child drowning deaths 2019-2025

by | May 6, 2026

Child drowning deaths – 1 April 2019 to 31 March 2025

Download the data tables

In partnership with

This summary report aims to describe the number and characteristics of child drowning deaths notified to Child Death Overview Panels in England, based on the latest available data from the NCMD. It builds on the findings of the NCMD thematic report Deaths of children and young people due to traumatic incidents and reinforces the recommendations made. 

The NCMD programme team wish to acknowledge that the death of each child is a devastating loss that profoundly affects bereaved parents as well as siblings, grandparents, extended family members, friends and professionals. They also wish to thank all the families who shared their data and experiences, and the Child Death Overview Panels who submit detailed evidence on every death to the database. 

In the year ending 31 March 2025, there were a further 31 child (0–17 years) drowning deaths notified to Child Death Overview Panels in England (Figure 1); 9 fewer deaths than in the previous year. This represents a death rate of 2.63 per 1,000,000 children aged 0-17 years. 

Across the six years from 1 April 2019 to 31 March 2025, the total number of drowning deaths was 196, representing an estimated death rate of 2.77 per 1,000,000 children per year.

Figure 1: Number of child (0-17 years) deaths due to drowning, from 1 April 2019 to 31 March 2025

 

  • In total, there were over twice as many drownings of male children (n=135) than female children (n=61). This represented a death rate of 1.77 per 1,000,000 per year for female children and 3.73 per 1,000,000 per year for male children. 
  • Children aged under 5 and those aged 13-17 years continued to be most at risk of drowning, with a death rate of 3.74 and 4.88 per 1,000,000 population per year, respectively. The death rate of 13–17-year-olds increased from 3.66 per 1,000,000 children per year in the 3-year period from April 2019 to March 2022, to 6.10 in the 3-year period from April 2022 to March 2025.  
  • Children of black ethnicity had more than three times the risk of drowning compared with children of white ethnicity (8.19 per 1,000,000 per year compared to 2.30 per 1,000,000 per year).  
  • The risk of drowning was more than twice as high for children who lived in the most deprived areas (3.72 per 1,000,000 per year) compared to the least deprived (1.67 per 1,000,000 per year). 
  • Death rates differed across regions with children living in the North West having the highest drowning rate of 4.05 per 1,000,000 population per year, and those living in the East of England having the lowest (1.75 per 1,000,000 population per year).
  • 24 children (12%) normally resident in England drowned abroad during the six-year period, 15 of whom drowned in a swimming pool. 
  • Deaths occurred across the year, with the largest proportion (48%, n=95) occurring during summer months (Jun-Aug). Deaths during spring months (Mar-May) increased from 15 deaths in the 3-year period from April 2019 to March 2022, to 24 deaths in the 3-year period from April 2022 to March 2025. Deaths during autumn months (Sep-Nov) increased from 7 deaths in the 3-year period from April 2019 to March 2022, to 21 deaths in the 3-year period from April 2022 to March 2025. 
  • Across the six years, 95 children (48%) drowned inland (in places such as rivers and lakes). This increased from 33 deaths between in the 3-year period from April 2019 to March 2022, to 62 deaths in the 3-year period from April 2022 to March 2025.   
  • 51 children drowned in the bath (26%) (26 from April 2019 to March 2022, and 25 from April 2022 to March 2025).  
  • Of 14 infants (children aged under 1 year) who drowned in the bath in the last 6 years, 9 were known to be using a bath seat, 2 were not using a bath seat, and there were 3 deaths where this information was not known. 

For deaths that had been reviewed by a CDOP (n=143):  

  • Where it was recorded (n=132), 116 (88%) of child drownings occurred in the absence of adult supervision. 
  • 71% (n=100/140) of child drownings involved a child who had current or previous social care involvement (for any reason). Where the information was known, household violence was reported in 39 deaths (29%), and abuse or neglect concerns were reported in 27 deaths (20%). 
  • 11 children (all ages) were diagnosed with autism or ADHD, and a further 9 were either awaiting assessment for, previously assessed for, or suspected of having autism or ADHD. These 20 deaths represent 14% of all drowning deaths.
  • CDOPs identified modifiable factors in 87% (n=124/143) of the reviews. Where information was available (n=94), the most common modifiable factors reported were lack of appropriate supervision (n=38), absent or non-visible warning signs (n=20), accessible water (n=19), the availability of safety equipment (n=17), and unsafe appliances or environment (n=14).

Methodology and limitations

Cohort identification

Deaths of children aged 0-17 years that occurred between 1 April 2019 and 31 March 2025 and were notified to NCMD were identified, where they were categorised as drowning using information recorded at notification. 

At the time of extraction, 82% (n=160/196) of the cases had been reviewed by a CDOP.

 

Drowning definition

Deaths where drowning was suspected or confirmed to be the cause of death at the point of notification were included for analysis. This is defined as where two or more clinicians coded the death as by drowning, based on information recorded within the notification form (or were included in the NCMD thematic report Deaths of children and young people due to traumatic incidents for the years 2020-2022). 

The definition includes deaths across a range of locations, including in residential settings (e.g., baths, hot tubs, ponds (domestic), private swimming or paddling pool), those in public swimming pools, and open water (e.g., sea, lake, quarry, reservoir, river, harbour, or canal). This includes deaths of children normally resident in England, where the drowning occurred outside of England.  

Deaths reported in tables 1-5 used all drownings where two or more clinicians coded the death as by drowning at notification (n=196). The deaths reported in tables 6-8 were further restricted to those that had reviewed by a CDOP before 31 March 2026, and where the primary category of death was recorded by the CDOP as ‘Trauma or other external factors’ (n=143). There were 36 deaths where the CDOP review was ongoing at 31 March 2026, and a further 17 deaths where the CDOP review was completed but the primary category of death was not recorded as ‘Trauma or other external factors’. These deaths were included in the analysis for tables 1-5 for consistency in the measurement of temporal trends, but excluded from tables 6-9. The information reported in these tables 6-9 is collected throughout the child death review process, therefore restricting this cohort presents the most complete data where the cause of death was confirmed as accidental drowning.  

 

Autism/ADHD – cohort identification

Children with autism or attention deficit hyperactivity disorder (ADHD) were identified using a text search on all the information contained within the NCMD. This was because the statutory data collection forms do not include a question around whether the child was diagnosed with autism or ADHD. A list of key words [‘autism’, ‘autistic’, ‘ASD’, ‘aspergers’, ‘ADHD’, ‘attention deficit hyperactivity disorder’] was derived and all records within the timeframe were searched. These deaths were then reviewed by the NCMD team to confirm that the child was diagnosed with autism or ADHD, and these deaths are reported in Table 8.  Any deaths that mentioned suspected diagnosis/traits of autism or ADHD, or where they were awaiting diagnosis, were recorded separately. 

 

Data extraction

The data extract used for deaths that occurred between 1 April 2019 and 31 March 2025 within this report was extracted on 19 February 2026. Following extraction, the data was validated throughout February and March 2026 before further analysis. This included a review of previous cases to assess where further data may have become available. In a limited number of cases, this may result in some differences to previously reported data, for example in relation to the presence of adult supervision.

 

Population estimates data

To calculate drowning rates, the ONS mid-year population estimates in 2021 for 0–17-year-olds were used as denominators. The data sources are reported under each table. The rate is presented per 1,000,000 children per year. The population was assumed to be stable throughout the six-year period and was not adjusted across the years.   

Limitations

This publication had limitations. This work is based on statutory data reported to NCMD, and previous work has shown good validation and coverage.  

The suspected category of death used to identify the deaths in this work (tables 1-5) is provisional, based on information available at 48 hours, and final underlying cause of death assigned by the CDOP may differ in some cases. 82% (n=160/196) of deaths had been reviewed by a CDOP, and 17 deaths had a primary category of death that was not recorded as ‘Trauma or other external factors’. This means that the data in tables 1-5 includes some deaths where the information was limited (as the review was ongoing) and also some deaths where the final category of death differed (e.g., ‘Deliberately inflicted injury, abuse, or neglect’).  

To ensure consistency across the whole period, this work did not include deaths where the final underlying cause of death assigned by the CDOP was drowning, but the suspected cause of death at the point of notification had not been identified as drowning, though this number is expected to be small. In addition, some changes in the methodology used to identify the cohort across the six years may impact the interpretation of temporal trends.  

The main limitation of this work is likely to be the precision of the estimates. With childhood death by drowning a rare event, small numbers make interpretation difficult.