Variable Number Of Rows In Macro

I'm trying to write a macro with a variable number or rows depending on the total number of rows the workbook has.

I tried like this:

Code:
Sub provaanova()
Application.Run "ATPVBAEN.XLAM!Regress", ActiveSheet.Range(("$J$2:$J$") & Range("J" & Rows.Count).End(xlUp).Row) _
, ActiveSheet.Range(("$J$2:$J$") & Range("J" & Rows.Count).End(xlUp).Row), False, False, 99, "ANOVA", False, _
False, False, False, , False
Cells.Select
Cells.EntireColumn.AutoFit
Range("A1").Select
End Sub

and:

Code:

Sub Provaregress()
r = ActiveSheet.Cells.SpecialCells(xlCellTypeLastCell).Row
Application.Run "ATPVBAEN.XLAM!Regress", ActiveSheet.Range("$J$2:$J$r") _
, ActiveSheet.Range("$K$2:$M$r"), False, False, 99, "ANOVA", False, _
False, False, False, , False
Cells.Select
Cells.EntireColumn.AutoFit
Range("A1").Select
End Sub

But it gives me an error. The only way i was able to do it without error was:

Code:
Sub Provaregress()
Application.Run "ATPVBAEN.XLAM!Regress", ActiveSheet.Range("$J$2:$J$53968") _
, ActiveSheet.Range("$K$2:$M$53968"), False, False, 99, "ANOVA", False, _
False, False, False, , False
Cells.Select
Cells.EntireColumn.AutoFit
Range("A1").Select
End Sub

But as u can see it has a fixed number of rows. ( I need variable number of rows detected on column J, and then run regression with K,L,M as depending variables).