For lek-breeding birds, lek attendance can be correlated with mating success. Variability in lek attendance could confound interpretation of male reproductive effort and complicate the use of lek counts as an index to monitor abundance. We assessed the daily probability of male Sage Grouse Centrocercus urophasianus lek attendance and explored implications of attendance on lek counts. We fitted 145 males with global positioning system (GPS) transmitters over 4 years in Carbon County, Wyoming. We evaluated influences of lek size and topography, date, weather, and bird characteristics such as age on daily morning lek attendance. The daily probability of attendance ranged considerably each year, from 0.120 ( x, 95% CI 0.051-0.259) in 2012 to 0.917 (95% CI 0.844-0.957) in 2013 with peak attendance dates ranging from 8 April (2012) to 11 May (2011). Attendance decreased with increasing precipitation on the observation day. Only 44-79% of lek counts occurred on days without precipitation and with high attendance (i.e. within 0.1 probability of peak predicted attendance). Although lek counts and population abundance, predicted using attendance rates, followed a similar trend, the relationship was not significantly correlated. We provide empirical evidence supporting current lek-count protocols: managers should avoid counting leks on days with precipitation because attendance is reduced. Although managers sometimes only complete one to two lek counts per year on active leks, completing at least three lek counts as recommended in protocols increases chances for higher male counts and improves the relationship between counts and abundance. Attendance varies annually, making it challenging to use lek counts to assess regional population trends over short time periods unless attendance is accounted for.