Introduction: Reducing unwanted variation is one of the goals of Norwegian health policy. We mapped variations among municipalities in Health Region West in the use of specialist health services that fall under the chapter on mental health care and substance use in ICD-10, and looked at various variables that may be related to these. Method: Data were retrieved from the Samhandlingsbarometer, an online open-source portal, and systematized by the authors. The annual rate per 1,000 inhabitants of hospital admissions and outpatient/day treatments per municipality was used as the result variables. Bivariate and multivariable linear regression were useful for investigating the association between independent variables and outcome variables. Results: The results showed large differences in the use of specialist health services in the municipalities, with a factor of 6.7 (hospital admissions) and 8.3 (outpatient clinic contacts). In adjusted analyses, we found that higher overall costs for health and care services in the municipalities were associated with fewer hospital admissions (β –.378 – 95% CI –.616; –.142) and fewer outpatient/day treatments (β –.541 –.649; –.323). The number of users in municipal mental health services was associated with more hospital admissions (β .514 .298; .726) and more use of psychotropic drugs in connection with outpatient/day treatments (β .485 .263; .703). Home-based care for younger inhabitants (< 67 years) was also positively associated with the use of both hospital admissions (β .494, .266; .724) and outpatient/day treatments (β .471 .239; .703). Both the prevalence of marginalisation (not in employment, education or training) and a high level of education were associated with higher use of outpatient clinics (β .355 .132; .577 and β .296 .064; .528), while a dispersed housing pattern was associated with fewer hospital admissions (β –.421 –.679; –.158) and outpatient day treatments (β –.386 –.627; –.145). The explained variance of the multivariable models was small to moderately high (R2 .07 – .51). Conclusion: Morbidity, use of resources for municipal health care, and sociodemographic differences are related to the rate of admissions and outpatient/day treatments. The findings do not unequivocally indicate that greater resource use in the municipalities covaries with less use of specialist health services, although some of the findings point in that direction. Relative inequalities in municipalities, where many inhabitants have a high level of education and where many experience marginalisation (not in employment, education or training) are associated with high consumption of specialist health services in mental health care. Keywords: practice variation, specialist health service, psychiatry, substance use treatment