Objective To characterize the communication ritual in the pharmacist–customer dyad during a blood pressure measurement and counselling episode. Setting A Portuguese urban community pharmacy. Method An exploratory, cross-sectional design was used. Participants were purposively selected and data collected via audio recording, as well as demographics and clinical information via questionnaire. Encounters’ verbal content was transcribed verbatim, utterances identified, time stamped, and classified according to a coding scheme of fifteen categories. All data was statistically analyzed using SPSSv17. Main outcome measures Four dialogue structures: speaker turn, interactivity, turn density and turn duration measurements. Results From a total of 51 participants, 72.5% were female with a median age of 66 years. The average systolic blood pressure was 140 mmHg, while the diastolic was 78 mmHg. The blood pressure measurement episode lasted for 5:35 min, with an average of 81 utterances. From all utterances registered, 55.3% were produced by the customer. Visits averaged 38 speaker turns, with an interactivity rate of approximately 7 turns per episode minute. For pharmacists, turn duration averaged 7.0 s and turn density 2.1 utterances. The customers’ turns comprised a mean of 8.0 s, with 2.4 utterances. Longer episodes were related to more speaker turns and greater customer turn density and duration, but lower dialogue interactivity. The interactivity rate was also lower when the customers’ utterances increased. Pharmacists asked more questions (essentially closed ones), while the customers gave more information. No significant associations were observed between elderly/non-elderly and gender in relation to all communication variables. However, an increased number of speaker turns and closed-questions were associated to a higher systolic pressure. Conclusion It seems that pharmacists tend to control the content of the dialogue, while customers have more influence on the visit duration and interactivity. Specific hypertensive episodes induce a higher information search. Since the closed questioning format was prevalent, it seems that open information exchange was limited, mainly serving confirmation purposes rather than having a true exploratory nature. Although talk dominance is balanced, further analysis is required to better inform these results, which would confirm the low interactivity and the reduced information-seeking behaviour showed in the counselling episodes.