Performance of query execution

Question

Hello friends,

Suppose I have 5000 records and I want to perform an operation on it.

I want to use 2 columns in sub-query for all above records.

Example:

Way-1:

DECLARE @TempSizeID INT;
SET @NewSizeID = ISNULL((SELECT ISNULL(id,0) FROM tb_sizes WHERE StoreID=@NewStoreID),0);

IF(@NewSizeID = 0)
BEGIN
    -- sub-query for insert data
    SET @NewSizeID = SCOPE_IDENTITY()
END
ELSE
BEGIN
    IF NOT EXISTS (SELECT ISNULL(SizeID,0) FROM tb_ProductColorSizeMapping WHERE ProductID = @NewProductID AND SizeID = @NewSizeID)
    BEGIN
        -- sub-query for insert mapping data
    END
END

Way-2:

DECLARE @TempSizeID INT;

IF(@ISNULL((SELECT ISNULL(id,0) FROM tb_sizes WHERE StoreID=@NewStoreID),0) = 0)
BEGIN
    -- sub-query for insert data
    SET @NewSizeID = SCOPE_IDENTITY()
END
ELSE
BEGIN
    IF NOT EXISTS (SELECT ISNULL(SizeID,0) FROM tb_ProductColorSizeMapping WHERE ProductID = @NewProductID AND SizeID =
    (ISNULL((SELECT ISNULL(id,0) FROM tb_sizes WHERE StoreID=@NewStoreID),0)))
    BEGIN
        -- sub-query for insert mapping data
    END
END

In way-1 it number of queries is less but it takes storage to store id
In way-2 it number of queries is more but and not store id
Can anybody suggest me which is fast?.