We developed the Geographical Proportional-to-size Street-Intercept Sampling (GPSIS) method in order to obtain a sample of nightlife-goers which accounted for the diversity of spaces, patrons and locations within two Swiss cities. Popular nightlife zones were identified and quantified using social media data and local experts' knowledge. Young people were recruited in the streets on Friday and Saturday nights on three consecutive weekends using the 'fixed-line method, pro-rated for the zone's estimated popularity. Of the 3092 young adults approached, 896 agreed to pre-register. The importance of recruitment in multiple zones and over multiple weekend-days was evidenced by significant variations in participant demographics and registration rates between recruitment zones, times and weather conditions. To conclude, by combining a geographical approach with "in situ" recruitment, GPSIS has considerable potential as a tool for recruiting samples that represent the diversity of the nightlife population and spaces.