The purposes of this cross-sectional study of 75 outpatients of a general oncology clinic were to assess the subjective and objective factors associated with dyspnea in cancer patients and to characterize factors that might contribute to respiratory muscle weakness demonstrated in a previous study. Patients with moderate to severe shortness of breath completed visual analogue scales (VAS) of shortness of breath (SOB) and anxiety; other data were acquired from pulmonary function tests, including maximum inspiratory pressure (MIP) and expiratory pressures; chest radiography; arterial blood gases; measurement of hemoglobin, serum potassium, phosphate, calcium, albumin, and magnesium; and ultrasound study of the diaphragm for thickness and excursion. The correlation coefficient between SOB VAS and anxiety VAS was 0.26 (P = 0.03). In stepwise multiple regression analyses, only the regression coefficient for anxiety remained significant at P < 0.05 in the multivariate model with SOB VAS as the dependent variable. The multivariate model using MIP (a measure of respiratory muscle strength) as the dependent variable, found significance for total diaphragmatic excursion, hemoglobin, phosphate, residual volume over total lung volume, vital capacity, percent predicted total lung capacity, oxygen saturation, and forced vital capacity. The regression coefficients for these variables were significant at P < 0.05 and the model accounted for 58% of the variance of MIP.