>Hi,
>
>I have to create a SQL Passthrough in which the WHERE clause filters if the days passed between a date in one record to the date in the previous record are greater than let say 10 days, within a specific grouping of records.

IF your rows have a consecutively numbered key column you might be able to use something like the following, which works in Fox but requires 3 SELECTs.

Your backend (SQL Server, Oracle, ??) may support more powerful SQL which may let you combine some of the SELECT steps. OTOH you'll probably have to do some sort of arithmetic and/or CAST()/CONVERT() with datetimes since backends typically don't support pure date types.

* Build and populate cursor for test:
CREATE CURSOR PrevTest ;
	(iID I, ;
	dDate D)

INSERT INTO PrevTest (iID, dDate) VALUES (1, {^2004-11-22})
INSERT INTO PrevTest (iID, dDate) VALUES (2, {^2004-11-23})
INSERT INTO PrevTest (iID, dDate) VALUES (3, {^2004-12-04})
INSERT INTO PrevTest (iID, dDate) VALUES (4, {^2004-12-06})
INSERT INTO PrevTest (iID, dDate) VALUES (5, {^2004-12-06})
INSERT INTO PrevTest (iID, dDate) VALUES (6, {^2004-12-18})
INSERT INTO PrevTest (iID, dDate) VALUES (7, {^2004-12-19})
INSERT INTO PrevTest (iID, dDate) VALUES (8, {^2004-12-31})
INSERT INTO PrevTest (iID, dDate) VALUES (9, {^2005-01-01})
INSERT INTO PrevTest (iID, dDate) VALUES (10, {^2005-01-21})

* Get 1st temporary cursor with "previous" row ID in each row:
SELECT ;
	*, ;
	iID - 1 AS iPrevID ;
	FROM PrevTest ;
	INTO CURSOR Temp1 ;
	NOFILTER

* Look up previous date for each row (if present),
* if not present use NVL() and a dummy date far in the future
* so the date difference will always be negative:
SELECT ;
	A.*, ;
	NVL(B.dDate, {^2005-12-31}) AS dPrevDate ;
	FROM Temp1 A ;
		LEFT OUTER JOIN PrevTest B ;
		ON B.iID = A.iPrevID ;
	INTO CURSOR Temp2 ;
	NOFILTER

* Now just get the rows that meet the date difference criterion:
SELECT ;
	* ;
	FROM Temp2 ;
	WHERE dDate - dPrevDate >= 10 ;
	INTO CURSOR Result

If you examine the cursors you'll see how it works.