Objective The aim of the study was to compare criterion validities of the WHO‐Five Well‐being Index (WHO‐5) and the Geriatric Depression Scale 15‐item version (GDS‐15) and 4‐item version (GDS‐4) as screening instruments for depression in nursing home residents. Methods Data from 92 residents aged 65–97 years without severe cognitive impairment (Mini Mental State Examination ≥15) were analysed. Criterion validities of the WHO‐5, the GDS‐15 and the GDS‐4 were assessed against diagnoses of major and minor depression provided by the Structured Clinical Interview for DSM‐IV. Subanalyses were performed for major and minor depression. Areas under the receiver operating curve (AUCs) as well as sensitivities and specificities at optimal cut‐off points were computed. Results Prevalence of depressive disorder was 28.3%. The AUC value of the WHO‐5 (0.90) was similar to that of the GDS‐15 (0.82). Sensitivity of the WHO‐5 (0.92) at its optimal cut‐off of ≤12 was significantly higher than that of the GDS‐15 (0.69) at its optimal cut‐off of ≥7. The WHO‐5 was equally sensitive for the subgroups of major and minor depression (0.92), whereas the GDS‐15 was sensitive only for major depression (0.85), but not for minor depression (0.54). For specificity, there was no significant difference between WHO‐5 (0.79) and GDS‐15 (0.88), but both instruments outperformed the GDS‐4 (0.53). Conclusions The WHO‐5 demonstrated high sensitivity for major and minor depression. Being shorter than the GDS‐15 and superior to the GDS‐4, the WHO‐5 is a promising screening tool that could help physicians improve low recognition rates of depression in nursing home residents. Copyright © 2013 John Wiley & Sons, Ltd.