Comprehensive population parameter data are useful for assessing effective conservation actions. The Fennoscandian arctic fox Alopex lagopus is critically endangered and the population size is estimated at 120 individuals that are fragmented into four isolated populations. Here, we use molecular tracking and visual observations to estimate population size and survival in one of the populations on the Swedish mountain tundra during a year of low food availability. We collected 98 arctic fox faecal samples during the winter of 2006 and recorded visual observations of ear-tagged individuals during the summer of 2005 and 2006. The faecal samples were analysed for variation in nine microsatellite loci and matched to the genetic profiles of previously ear-tagged individuals from 2001 to 2005. During winter 2006, the minimum number alive was 12 individuals using visual observations, 30 using molecular tracking and 36 by combining the datasets. Population size was estimated through mark–recapture for the molecular tracking and visual observation datasets and through rarefaction analyses for molecular tracking data. The mark–recapture estimate for visual observations was uninformative due to the large confidence interval (CI) (i.e. 6–212 individuals). Based on the molecular tracking dataset combined with the minimum number alive for visual observations and molecular tracking, we concluded a consensus population size of 36–55 individuals. We also estimated the age-specific finite survival rate during 1 year (July 2005 to July 2006) by combining molecular tracking with visual observations. Juvenile survival on a yearly basis was 0.08 (95% CI 0.02–0.18) while adults had a survival of 0.59 (95% CI 0.39–0.82). Juveniles displayed a lower survival than the adults during autumn (P<0.01) whereas no age-specific survival difference during spring was found. The risk of negative effects due to the small population size and low juvenile survival is accordingly considerable.