Since you mention SMT lands, I will assume we are talking about SMT solder balls, not wave. In a word, yes, it can have an effect. The solder mask typically has an under-cut edge that may trap paste. Ideally we try to prevent paste from ever getting there, by:

• Designing the land and aperture to ensure we don't over-print the edge of the land
• Maintaining tight control in the printing process, including providing adequate support in all areas, ensuring registration accuracy, optimizing/controlling printing parameters (squeegee pressure and angle, speed), and controlling paste properties (temp/humidity of environment, re-use, open time...).
• Controlling reflow process to minimize slump

When all of the above are controlled, the only solder balls that usually still appear are related to component-paste interactions, and these do not appear in the pad-to-mask gap but along or under component bodies.

One component type where optimization of the mask-to-land relationship is very important is for fine-pitch leaded components (QFP, TSOP, etc.). It's best practice to remove any mask web from between apertures, and to use the minimum gap required by mask registration variability at the land ends.

Paste apertures should be reduced slightly in length to avoid over-printing the ends at maximum misregistration. Typically, under-sizing by 0.002" per end is enough.

F. B.