Original article ISPreview UK:Read More
A new study from Broadband Genie has analysed 145,926 internet speed tests in order to identify the top slowest and fastest ten UK streets for broadband. The slowest street was found to be Heol-Y-Fedw (Cymmer) in Wales on an average download of just 0.81Mbps (Megabits per second), while the fastest was Inglewood Avenue in Derby (England), where residents enjoy speeds of 1.21Gbps (1,210Mbps).
According to Thinkbroadband’s latest independent modelling, some 89.7% of UK premises are currently within reach of a gigabit-capable fixed broadband network (1000Mbps+), which primary reflects the combination of full fibre (FTTP/B) and hybrid fibre coax (HFC) lines (the latter is largely only from Virgin Media). Similarly, some 98.45% of premises should be within reach of a fixed “superfast broadband” (30Mbps+) network.
However, the new study is based on consumer speedtests, rather than network availability. One issue here is that such reports aren’t able to accurately reflect the underlying availability of faster networks and are thus more a reflection of localised consumer take-up (i.e. results from people still on slower networks/packages may weight against those on faster ones in the same street).
The new report does acknowledge the aforementioned issue and highlights how occupants on the slowest listed streets all have access to Fibre-to-the-Cabinet (FTTC / VDSL / SOGEA) based broadband lines, which can provide a download speeds of around 35Mbps to many of them. In addition, seven of the slowest streets had access to “ultrafast” speeds (100Mbps+) and four of the slowest were even covered by gigabit full fibre (FTTP) networks.
Clearly, issues of awareness, demand / desire and the fear of switching to a new service or ISP may still be holding some consumers back from making the change to a faster broadband connection. This could, in some cases, also mean they’re stuck paying higher prices for a legacy connection that underperforms.
However, speedtest-based studies like this can also be influenced by other factors, such as poor home wiring (copper telecoms lines), local (home) network congestion and slow WiFi etc. In short, take these results with a pinch of salt and remember, this study won’t be reflecting every street in the UK because a minimum of three speed tests from three unique IP addresses via a commercial fixed line broadband ISP were required for inclusion. At least 8 residential properties were also required at a postcode, so quite a lot of streets will be missed.
Similarly, the fastest streets are usually those covered by the handful of providers capable of offering them speeds in the 7-10Gbps range, such as B4RN, Youfibre and a few others.
Slowest 10 streets for UK broadband (2026)
| Rank | Street | Download speed (Mbps) |
| 1 | Heol-Y-Fedw, Cymmer, Port Talbot | 0.81 |
| 2 | Turnberry Crescent, Bridge of Don, Aberdeen | 1.06 |
| 3 | Wesley Street, Maesteg | 1.45 |
| 4 | Occupation Lane, Broadholme, Lincoln | 1.63 |
| 5 | Rossiter Road, London | 1.74 |
| 6 | Quarry Close, Handbridge, Chester | 2.23 |
| 7 | Langley Street, Langley | 2.23 |
| 8 | Jessop Road, Rogerstone, Newport | 2.62 |
| 9 | Wakefield Close, Hurley, Atherstone | 2.66 |
| 10 | Rheolau Terrace, Pontypridd | 2.80 |
Fastest 10 streets for UK broadband (2026)
| Rank | Street | Download speed (Mbps) |
| 1 | Inglewood Avenue, Derby | 1.210Gb |
| 2 | Moatview Park, Dundonald, Belfast | 1.146Gb |
| 3 | Reynolds Avenue, Romford | 1.014Gb |
| 4 | Sarum Close, Salisbury | 1.009Gb |
| 5 | Broad Lane, Wolverhampton | 947.1Mb |
| 6 | Limbury Road, Luton | 946.8Mb |
| 7 | Baberton Mains Drive, Edinburgh | 943.2Mb |
| 8 | Park Road, Camberley | 931.3Mb |
| 9 | Bramble Drive, Westbury | 928.4Mb |
| 10 | Powerscourt Road, Portsmouth | 926.3Mb |
The government’s Project Gigabit scheme is currently working to help extend broadband ISP networks capable of delivering download speeds of at least 1000Mbps (1Gbps) to achieve “nationwide” coverage (c.99%) by 2030 2032 (here) – focusing on the commercially unviable areas (usually rural and semi-rural locations). But the impact of this won’t fully show in the results above until everybody moves over to such lines.