Postoperative speech function may be influenced by a number of treatment variables. The objective of this study was to examine the relationships among various treatment factors to determine the impact of these measures on speech function. Speech function was tested prospectively in 142 patients with surgically treated oral and oropharyngeal cancer 3 months after treatment. Each patient's speech was recorded during a 6- to 7-minute conversation and while performing a standard articulation test, producing speech outcome measures of percent correct consonant phonemes and percent conversational understandability. Correlational analyses were used to determine the relationships among the speech outcome measures and 14 treatment parameters. Speech function was mildly to moderately negatively correlated with most surgical resection variables, indicating that larger amounts of tissue resected were associated with worse speech function. Overall measures of conversational understandability and percent correct consonant phonemes were related to extent of oral tongue resection, floor of mouth resection, soft palate resection, and total volume of tissue resected. These relationships varied depending on the method of surgical closure. Method of surgical reconstruction had a profound impact on postoperative speech function 3 months after treatment and was an important factor in determining how oral tongue resection influenced articulation and intelligibility. The combination of closure type, percent oral tongue resected, and percent soft palate resected had the strongest relationship with overall speech function for patients with surgically treated oral and oropharyngeal cancer 3 months after treatment.