AbstractThis study investigated the fine‐scale land parcel features and their influences on the intra‐surface urban heat island (SUHI) of four dense urban functional zones in downtown Shanghai. A thermal sharpening method was employed to map the high‐resolution intra‐SUHI intensity (SUHII) and magnitude (SUHIM) by integrating the cokriging interpolation of Landsat 8 thermal band 10 and Quickbird datasets. Together with three area proportions (area proportion of the cooling surface/CoolAP, area proportion of the buildings/BuildAP, and area proportion of the other impervious surface/OtherAP), the mean distance between the nearest neighboring buildings (BuildDist) and the building average story number (BuildSN) were used to characterize the land parcel features. The results showed that both the composition and the design attributes of land parcels were significantly related to the intra‐SUHI indicators (SUHII and SUHIM). Additionally, remarkable seasonal variations existed in the intra‐SUHI indicators. The results of the partial least square regression models explained 64.6–65.2% of the variance of the SUHIIs and 63.0–63.9% of the variance of the SUHIMs on hot days and 11.70–18.90% of the variance of the SUHIIs and 11.10–15.90% of the variance of the SUHIMs on cool and chilly days, respectively. This study provided compelling evidence that for a given land parcel, when fixing the effects of BuildAP or OtherAP, high levels of CoolAP, BuildSN, and BuildDist will help create a cooling microclimate on hot days. In contrast, high‐density land parcels with sparse cooling surface exhibit much higher SUHII and SUHIM levels than do low‐density parcels with dominant cooling surface.