Advancements in wearable sensor technologies have revolutionized healthcare, particularly in nurses stress level detection using accelerometer-based data. Accurate detection of nurse stress throughout a day holds significant potential to enhance healthcare outcomes and optimize nursing workflows. However, existing research in this domain has been limited by small data-sets and short sampling duration’s, leading to constrained generalizability. To address these limitations comprehensively, this paper presents a data on nurses’ mental health in a practical setting are gathered using a wearable accelerometer sensor. To record the physiological and behavioral traits of the nurses, a single triaxial accelerometer is mounted on their heads. Multiple machine learning methods, such as Nearest Neighbor Classifiers (k-NN), Neural Networks (NN), Support Vector Machines (SVM), Naive Bayes (NB), Discriminant Analysis (DA), Decision Trees (DT), and Ensemble Classifiers, were used to assess and compare the accuracy of nurses’ stress levels. The Medium Gaussian-SVM study showed the best average performance of 80.4% and mean efficacy F-measure of 80.2% across all cases. whereas the Coarse Gaussian-SVM evaluation exhibited the lowest mean performance of 26.5%, accompanied by a mean F-measure of 26.1%. On a scale of 1–7, a questionnaire evaluated privacy issues. The results show our method’s potential for accurately determining nurses’ levels of stress, making it helpful for implementation in the future.