Analysis of hair cortisol concentrations (HCCs) is a promising method for monitoring long-term stress in mammals. However, previous measurements of HCCs in polar bears (Ursus maritimus) have yielded highly variable results, which are likely due to different methodological approaches. In this study, hair samples of zoo-housed polar bears were analyzed for cortisol with two independent immunoassays [an enzyme-linked immunoassay (EIA) and a chemiluminescence assay (CLIA)] and liquid chromatography–tandem mass spectrometry (LC–MS/MS). HCC measurements depended significantly on assay type applied, sample processing (cutting vs. powdering hair) and their interaction. Best agreement was observed between LC–MS/MS and CLIA (R² = 0.81 for powdered hair) and sample processing had a minor, albeit significant, effect on obtained HCC measurements in these assays (R² > 0.9). EIA measurements were consistently higher than with the other assays. HCC measurement was validated biologically for CLIA and LC–MS/MS in one male polar bear that experienced considerable stress for a prolonged period of time (> 18 weeks). Subsequently, by using the validated LC–MS/MS the measurement of cortisol could be complemented by the analysis of other steroids including cortisone, testosterone and progesterone levels from hair samples collected over a 9-month period (5–13 months) from six zoo-housed polar bears (five males, one female). No seasonal steroid variation was observed except in male progesterone levels. For all steroids except cortisone, a strong body region effect (neck or paw) was observed. Cortisol and cortisone, as well as progesterone and testosterone, concentrations were positively correlated. We show that hair steroid concentrations can be used to longitudinally measure stress and reproductive hormone axes in polar bears. The data established herein provide important basic information regarding methodology and study design for assessing hair steroid hormones.